With the rapid development of broadband services, interconnection between networks is used more and more to carry more services by the networks. One or more links and one or more nodes thereon in a network may be protected by various methods according to a technology applied by the network. With growing demands and increasing requirements for traffic protection, some operators also need to protect network interconnection. Preferably, the protection herein may be implemented by port aggregation. A common method may be port aggregation or link aggregation, and may be also loop protection. In link aggregation, a common method at present is port aggregation on one node, thus link aggregation can be only applied to link protection and protection required for a node on a network edge interface can be hardly implemented through the current link aggregation technology.
Therefore, it is proposed by the Institute of Electrical and Electronics Engineers (IEEE) standards organization to extend link aggregation to meet a requirement of increasingly diverse networking modes of a network interconnect area, and implement not only protection for a link, but also protection for an edge node. That is, a demand for network interconnecting protection with dual redundancy of links and nodes is implemented through a distributed Link Aggregation Group (LAG), i.e. a Distributed Resilient Network Interconnect (DRNI). That is, a portal of an aggregation group is formed by a plurality of nodes, and links formed by aggregating the plurality of nodes are aggregated to form an LAG. As shown in FIG. 1, two portals A and B of an LAG are both provided with two nodes. Portal A corresponds to System 1 and System 2 and Portal B corresponds to System 3 and System 4. A plurality of links of these four systems are aggregated to form an LAG. Dual protection of links and nodes may be implemented through such a distributed LAG. Here, System 1 and System 2 in Portal A communicate through an internal link, while System 3 and System 4 of Portal B also communicate through an internal link.
Currently, the DRNI implements protection and sharing for a load through distinguishing message services. However, it can be hardly ensured that traffic of the same service can be transmitted on the same aggregation link, because the traffic needs to be distributed by two ends respectively. The method for distributing traffic is contradictory to application of the DRNI in protection of a network interface and inconsistent with a demand of the Metro Ethernet Forum (MEF) for protecting a network interface. For protection, it is expected that a packet switching network is able to simulate Time Division Multiplex (TDM) to realize a kind of point-to-point service protection, and services affected by switching on a fault can be reduced as many as possible, i.e. switching of services is affected as little as possible.
Therefore, a feasible and effective method needs to be found in order to ensure that the same service is bidirectional on the same path when transmitted on DRNI.
At present, there is no effective solution for the problem above.